150k Tokens Deep: Why Context Recall Beats Compaction by up to 8x
dacorvo
• • 2NEW Articles from Team or Enterprise organizations will get promoted to the main section.
Introducing Olmo-core 3: Open, scalable training infrastructure for large MoEs
allenai
• • 22What a million datasets on Hugging Face tell us about open AI data
elrobert9
• • 2Is Space Bunny a MiniMax Model? Tokenizer Evidence and a Falsifiable Prediction
karmen-beatapi
• How to Test Space Bunny as a Coding Agent: Three Reproducible Tasks
karmen-beatapi
• Your Professor's AI Is Cheating on You: Why Pedagogy Requires Uncensored Tools
thevasilis
• • 1Your University's AI Is Running on Someone Else's Morals: Institutional Morality and EdTech
thevasilis
• • 1A $600 VM Beat the Frontier Labs: Full Results From Our Reproducible Aristotelian Benchmark
thevasilis
• • 1What a Point of Accuracy Costs
skylord
• The Nature of Consciousness
pet-rank
• • 1Why AI Providers Shouldn’t Grade Their Own Bills
bharathkoneti1225
• • 1Your Model Works in a Notebook. Now What? A Friendly Intro to NVIDIA Triton Inference Server
omarcevi
• AIAMR-L160 coauthor attribution probe
agentamrorg
• Hotel Lobby AI: What the Viral Trend Is and How to Access It
liliruli
• GPT IMAGE 2 in practice：Create Transparent PNGs and Verify the Exported Result
VulcanFeynman
• Qwen3.8-27B-pi: Effort-Ordered Reasoning for Agentic Coding
bytkim
• • 13Write an Eval That Can Say No
ericmey
• Ten thousand synthetic policyholders: a practical guide for actuaries, economists and social scientists
alidenewade
• Training a Looped Transformer: When to Repeat, Review, and Stop
TiGa-RCE
• Introducing AURORA: A Multilingual Leaderboard for Frontier Models on Non-English Agentic Tasks
Lilt-org
•